New Customer Value on Google Ads: How We Shifted Delphin Pêche’s Ad Spend Toward Pure Acquisition
Observation, Customer Match, LTV by market, value boost. A case study on the Google Ads account of Delphin Pêche, a fishing e-commerce retailer, and what you need to know before comparing time periods.
The Context: A Healthy Account, but “Polluted” Acquisition
Delphin Pêche is a fast-growing e-commerce retailer specializing in fishing equipment, operating across multiple European markets. Its strategy is clear: expand its customer base. On Google Ads, the account structure was clean, with Brand campaigns on one side and Non-Brand campaigns on the other.
This split isolates acquisition, but only partially. Generic keywords also capture existing customers who search for a product without typing the brand name. Consequently, reporting shows acquisition numbers that, in part, are not genuine new acquisitions.
To Google Ads, a first-time order and a tenth order hold the exact same value. The algorithm optimizes toward what converts easiest, and a customer who already knows the store converts much more easily than a stranger.
The Three Google Settings Not to Confuse
According to Google’s documentation on Customer Lifecycle Goals:
- No setting: All customers are treated equally.
- New customer only: The campaign targets new customers exclusively.
- Bid higher for new customers (“New Customer Value”): Bids are weighted more heavily for a new customer without excluding existing ones.
Our choice: The third option. Delphin wanted to drive new customer acquisition without losing loyal buyers or cutting off orders that remain profitable.
The Five-Step Method
Step 1: Observe first
Before adjusting a single bid, we added customer audiences to the account in “Observation” mode.
This phase revealed that a substantial number of existing customers were reordering via generic keywords. Campaigns designed for prospecting were actually bringing back returning customers. This data-backed evidence, taken directly from the account, justified the project.
Step 2: Leverage the client’s CRM data
Delphin possesses rich CRM data. Overall LTV and LTV by product category are calculated on the client side and vary by market. We didn’t reinvent these rules; we adopted them and translated them into advertising signals.
Step 3: Deploy Customer Match
Existing customers are uploaded via Customer Match so Google can identify them. Google recommends defining existing customer lists, ideally by combining Customer Match with website-based remarketing lists.
Step 4: Add value to every first purchase
The added value is calculated beforehand and adjusted based on product category. Certain gateway categories bring in customers who are highly likely to reorder later, meaning their first order carries greater lifetime value.
Step 5: Roll out in April, learn, adjust
The setup launched in April, followed by a learning phase for the smart bidding strategies, and then continuous adjustments.
Note: For more reliable classification, Google also recommends passing customer acquisition status directly through the conversion tag.
The Pitfall: The Dual Conversion Value
Once the boost is activated, two distinct columns coexist in Google Ads:
| Column | Content | Use case |
|---|---|---|
| Conversion value | Value including the new customer boost | Bid optimization / Smart Bidding |
| Original conversion value | Value excluding the boost | Reporting and performance comparisons (per Google docs) |
Google uses the example of a €100 order with a €30 adjustment: the algorithm optimizes for €130, but actual revenue remains €100. The boost is not actual revenue; it is merely a signal fed into the algorithm.
Impact for media buyers: Reports relying on the standard Conversion value column are artificially inflated. Comparing a period running this setup against a historical period without it makes no sense. You must use Original conversion value. Because this column is not always available in third-party reporting tools, automation can become tricky. It is a genuine constraint, but an essential one to manage.
Account Performance for Delphin Pêche
Weight of the Boost
Prior to April, the boost was virtually zero. Since rollout, it adds roughly 25% on top of real revenue, remaining fairly consistent month over month.
Year-over-Year Comparison
Comparing full months post-implementation against the same period last year, based on real value:
- Spend increased by ~20%.
- Real value (revenue) increased by ~25%.
- Order volume grew much faster, up by ~70%.
- Real ROAS remained stable.
- Boosted ROAS would suggest an increase of ~30%, which is misleading.
How to Interpret These Figures
- The reporting trap is confirmed: looking at the boosted column, one would assume a massive ROAS jump. In real terms, ROAS held steady.
- Growth is primarily driven by order volume alongside a lower average order value (AOV). This aligns perfectly with acquiring new customers via entry-level products, matching Delphin’s core growth strategy.
Methodological note: This represents correlation, not definitive causality. Other account variables evolved during this timeframe (seasonality, campaign structure, product catalog).
Key Takeaways
- Observe before taking action. The observation phase exposed the underlying issue and provided the data needed to secure buy-in.
- A conversion does not have a single fixed value. Bidding must reflect business goals, not just immediate basket size.
- CRM data is an advertising asset. Leveraged properly, it directly guides algorithmic bidding.
- Altering conversion values impacts reporting. You must plan ahead, document the change, and educate internal teams.
- Compare like with like. Always rely on Original conversion value for historical comparisons.
- LTV frameworks belong to the client and naturally differ across markets and product categories.
- Evaluate multiple KPIs together. ROAS alone won’t tell you if you are acquiring new customers; track order volume and AOV in tandem.
What to Avoid
- Comparing year-over-year performance using boosted conversion values.
- Trying to optimize everything around complex LTV models from day one. Start simple, measure the impact, then refine.
- Assuming “New Customer Value” excludes returning buyers. It does not; only “New Customer Only” mode does that.
- Skipping the observation phase. Without it, you cannot measure how much existing customers dilute your prospecting campaigns.
- Forgetting to inform reporting teams. Without proper context, the surge in conversion value will be misinterpreted.
Conclusion
For Delphin Pêche, separating Brand and Non-Brand campaigns was no longer enough. Segmenting campaigns is only half the battle: you must also segment your conversions. Assigning strategic value to a first order successfully directs ad spend toward pure acquisition, as long as performance is read using the right column.
